238 PROF. W. A. HASWELL : 



In front each germinal band divides into five segments, the first smaller 

 than the rest, which are subequal. Sections show that, like the germinal 

 bands, each segment is composed of two layers, a superficial, ectodermal, and 

 a deeper, mesodermal (PI. 18. fig. 63). 



The area between the two rows of somites continues to be covered dorsally 

 by a very thin ectoderm. Ventrally, on the other hand, at the stage when 

 segmentation has been completed, is a thicker layer of closely-set small cells 

 divided along the mid-ventral line by a raphe or suture (PI. 18. figs. 62 & 64). 

 This is obviously the rudiment of the ventral nerve-cord. Whether it is 

 formed from the thin ectodermal layer originally present in this position or 

 by lateral ingrowth from the two germinal bands has not been actually 

 determined. But the latter mode of origin, especially in view of the presence 

 of the median raphe, is by far the more probable. At this stage the rudi- 

 ment of the nerve-cord presents no trace of segmentation. 



The dorsal plate has thickened materially and extended over the anterior 

 (free) end of the embryo (PI. 18. figs. 58-64). The stomodrcum has extended 

 and assumed a rounded form. Posteriorly it gives off an extension — the 

 rudiment of the proventriculus — the original stomodjcum becoming the 

 pharynx. The epithelium of the former has formed around it a simple 

 layer of cells which tend to elongate radially, these are destined to form the 

 radial muscular fibres of the proventriculus. External to this and closely 

 applied to it are flattened cells which may form a continuous layer. 



The next change of importance is the appearance of the first rudiments 

 of the tentacles (figs. 64 & 65). The free or head end of the embryo had 

 already developed a solid thickening — the rudiment of the closely united 

 palpi. On the dorsal side of this appear two, to which a third (median) is 

 soon added, small knobs — the beginnings of the cephalic tentacles. About 

 the same time appear the rudiments of the anal cirri. The peristomial 

 tentacles do not appear till considerably later. At this stage there are six 

 pairs of mesodermal somites, the members of each pair widely separated from 

 one another. 



When the rudiments of the parapodia first appear as slight lateral pro- 

 jections they are developed from the somites. 



It appears to be clear that the ccelom cannot be formed in the same way 

 as in Polychfeta in general. At the stage when rudimentary seta? first 

 make their appearance in the parapodia the stomoda?um ends behind in a 

 mass of tissue in which the large yolk-granules are embedded. This tissue, 

 derived as has been seen, from the macromeres, no longer retains the 

 character of distinct cells, but has all the appearance of such a syncytium as 

 that presented by the enteric or endodermal mass of the " Acoelous " 

 Turbellaria or certain of the Rhabdocoeles. Jn these the food is digested in 

 larger or smaller vacuoles in the syncytium. In the embyro of Exogone 

 fustifera the yolk-granules, which take the place of food, are similarly 



